This is a repost of the pci-sysfs-rom-22.patch. No one has made any
comments on this version. All previous objections have been addressed.
Any objections to sending it upstream?
=====
Jon Smirl
[email protected]
_______________________________
Do you Yahoo!?
Win 1 of 4,000 free domain names from Yahoo! Enter now.
http://promotions.yahoo.com/goldrush
On Thursday, September 2, 2004 6:40 pm, Jon Smirl wrote:
> This is a repost of the pci-sysfs-rom-22.patch. No one has made any
> comments on this version. All previous objections have been addressed.
> Any objections to sending it upstream?
Hm, the last one I tried worked fine, but this one makes my qla card stop
working, but not right way. The system gets to init and then falls over,
maybe when it starts doing writes? The last version I tried seems to work ok
though. Has something changed in the PCI layer that would affect this?
Thanks,
Jesse
On Friday, September 3, 2004 10:27 am, Jesse Barnes wrote:
> On Thursday, September 2, 2004 6:40 pm, Jon Smirl wrote:
> > This is a repost of the pci-sysfs-rom-22.patch. No one has made any
> > comments on this version. All previous objections have been addressed.
> > Any objections to sending it upstream?
>
> Hm, the last one I tried worked fine, but this one makes my qla card stop
> working, but not right way. The system gets to init and then falls over,
> maybe when it starts doing writes? The last version I tried seems to work
> ok though. Has something changed in the PCI layer that would affect this?
It looks like hald is reading the rom attribute at boot, which either disables
decode for the qla mmio registers or otherwise panics. Bad hald.
Jesse
On Friday, September 3, 2004 10:45 am, Jesse Barnes wrote:
> On Friday, September 3, 2004 10:27 am, Jesse Barnes wrote:
> > On Thursday, September 2, 2004 6:40 pm, Jon Smirl wrote:
> > > This is a repost of the pci-sysfs-rom-22.patch. No one has made any
> > > comments on this version. All previous objections have been addressed.
> > > Any objections to sending it upstream?
> >
> > Hm, the last one I tried worked fine, but this one makes my qla card stop
> > working, but not right way. The system gets to init and then falls over,
> > maybe when it starts doing writes? The last version I tried seems to
> > work ok though. Has something changed in the PCI layer that would affect
> > this?
>
> It looks like hald is reading the rom attribute at boot, which either
> disables decode for the qla mmio registers or otherwise panics. Bad hald.
After disabling hald, things are ok. I can read ROMs that have been mapped
correctly. The problem was that sn2 doesn't map all ROMs by default, so the
qla card, which has a ROM, had it's resource structure filled with the value
in its BARs, which was of course not a usable CPU address. I can fix that
for sn2, so the patch is ok with me.
Jesse